Inline Memory Meets Reusable Skills: Memory-centric Framework for Vision-Language-Action Model

arXiv:2609.39794v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have shown strong promise for general-purpose robotic manipulation, yet adapting them to new tasks and domains remains inefficient: existing methods often rely on parameter tuning, incurring substantial costs and risking catastrophic forgetting of previously learned tasks. To address this, we propose extbf{Optimus-R}, a memory-centric VLA framework that formulates robotic adaptation as explicit query-skill memory tuning. Optimus-R introduces: (i) An extbf{Inline Memory Interface for skill extraction}. It inserts learnable memory tokens into the VLA prefix stream, allowing the backbone to derive control-aware query and skill representations within the native action-conditioning pathway. (ii) A extbf{Query-Skill Memory Bank for skill learning}. It externalizes skills into query prototypes for deciding what to retrieve and skill values for specifying how to act, supporting skill reuse and expansion with limited parameter updates. (iii) A lightweight extbf{Bridge-and-Adapt mechanism for skill updating}. It aligns target-domain queries and skills with the existing memory space through a lightweight adapter and residual memory updates. Experiments on in-domain adaptation, cross-domain adaptation, and lifelong learning show that Optimus-R enables data-efficient skill learning while mitigating catastrophic forgetting.

Source: arXiv cs.CV | 2026-10-01

Related